Despite more than a century of neuropsychological research, there is still disagreement about whether reading and spelling rely on independent or shared cognitive mechanisms and neural substrates. The current application seeks to improve our understanding of the relationship between reading and spelling in both cognitive and neuroanatomical terms. To achieve our objectives, we propose to conduct a detailed evaluation of written and spoken language function in a large group of individuals with left-hemisphere cortical stroke or semantic dementia and correlate behavioral performance with precise neuroanatomical information about lesion location. Specifically, we plan to document both associations and dissociations between the alexia and agraphia profiles of study participants and elucidate the critical cognitive and lesion variables responsible for the different patterns identified. These observations will provide the empirical database for testing the main prediction of shared-component models of written language processing, according to which damage to orthographic, phonological, and semantic representations should have similar consequences for reading and spelling. To maximize the reliability of the proposed lesion-deficit studies, we will obtain high resolution MRI scans and implement novel lesion analysis methods to correlate tissue damage with relevant behavioral measures of reading/spelling performance. Lesion data will also be correlated with composite measures of semantic, phonological, and visual/orthographic ability. In addition, we will use converging evidence from lesion studies in neurological patients with alexia/agraphia and functional imaging studies of reading/spelling in normal individuals to identify the neural systems involved in written language processing and test specific predictions regarding the role of distinct cortical regions in visual analysis, orthographic processing, phonology, and semantics. We expect that this large-scale investigation will produce new information about the cognitive and lesion variables that influence written language performance in individuals with acquired alexia/agraphia. Our prediction is that the behavioral and neuroanatomical data will support the hypothesis that reading and spelling rely on shared cognitive representations and neural systems within the central domains of orthography, phonology, and semantics. These findings should have important implications for theoretical models of written language processing. [unreadable] [unreadable] [unreadable]